Extract structured data from PDF
10× faster PDF parsing with layout analysis
Trained on structure, not images. CPU-only